Better sparse constructors

src/tensors/sparseblocktensor.jl

@@ -98,7 +98,15 @@ end
 Construct a sparse blocktensor with entries compatible with type `T` and space `W`.
 By default, the tensor will be empty, but nonzero entries can be specified by passing a tuple of indices `nonzero_inds`.
 """
-spzeros(W::TensorMapSumSpace) = spzeros(Float64, W)
+spzeros(W::TensorMapSumSpace, args...) = spzeros(Float64, W, args...)
+function spzeros(T::Type, cod::TensorSumSpace, dom::TensorSumSpace=one(cod), args...)
+    return spzeros(T, cod ← dom, args...)
+end
+function spzeros(
+    T::Type, cod::TensorSumSpace, nonzero_inds::AbstractVector{<:CartesianIndex}
+)
+    return spzeros(T, cod, one(cod), nonzero_inds)
+end
 function spzeros(
     ::Type{T}, W::TensorMapSumSpace, nonzero_inds=CartesianIndex{length(W)}[]
 ) where {T}
@@ -116,9 +124,17 @@ end
 Construct a sparse blocktensor with entries compatible with type `T` and space `W`.
 Each entry is nonzero with probability `p`.
 """
-sprand(V::VectorSpace, p::Real) = sprand(Random.default_rng(), Float64, V, p)
-sprand(rng::Random.AbstractRNG, V::VectorSpace, p::Real) = sprand(rng, Float64, V, p)
-sprand(T::Type, V::VectorSpace, p::Real) = sprand(Random.default_rng(), T, V, p)
+sprand(V::TensorMapSumSpace, p::Real) = sprand(Random.default_rng(), Float64, V, p)
+sprand(rng::Random.AbstractRNG, V::TensorMapSumSpace, p::Real) = sprand(rng, Float64, V, p)
+sprand(T::Type, V::TensorMapSumSpace, p::Real) = sprand(Random.default_rng(), T, V, p)
+sprand(V::TensorSumSpace, p::Real) = sprand(Random.default_rng(), Float64, V ← one(V), p)
+function sprand(rng::Random.AbstractRNG, V::TensorSumSpace, p::Real)
+    return sprand(rng, Float64, V ← one(V), p)
+end
+sprand(T::Type, V::TensorSumSpace, p::Real) = sprand(Random.default_rng(), T, V ← one(V), p)
+function sprand(rng::Random.AbstractRNG, T::Type, V::TensorSumSpace, p::Real)
+    return sprand(rng, T, V ← one(V), p)
+end
 function sprand(
     rng::Random.AbstractRNG, ::Type{T}, V::TensorMapSumSpace, p::Real
 ) where {T<:Number}